

210 NUMERICAL INCREASE OF PARTS OF PLAXTS. 



I 



it is frequent in foliar organs — as, abnormally in the fronds of Sco- 



lopendriuTrij or normally in the stamens of Tilia, Malva^ and some 



of the Amentacece^ and apparently also in Adoxa. In these androecia | 



the anther-lobes are separated by " fission," which is carried down 



more or less towards the base of the filament, and are unilocular. 



The words " chorisis " and " fission " are inappropriate ; for they 

 express only the appearance, and not the real developmental 

 process, as the stamens are not at first entire and then subse- 

 quently split ; and I would, were it possible, use only " bifurca- 

 cation," which simply describes the condition. Moreover, if cho- 

 risis be retained, I would limit its use to bifurcation alone. 



Whenever supernumerary anthers are two-celled, such a stamen 

 would not come under chorisis but hypertrophy, as in the case of 

 Hypericum. 



V. Atrophy. — To this cause I attribute all incised, cut, or 

 laciniated foliar organs, and thence also compound leaves, which 

 represent further conditions arrived at by the total arrest of the 

 parenchymatous tissue between the fibro-vascular bundles. Several 

 plants furnish transitional conditions which clearly prove this, as 

 Rubus fruticosuSj species of Cle?natis, Rhus TieteropTiylla^ &c. 



The development of alobed organ may easily give rise to a false 

 conception ; for the lobes begin as lateral outgrowths, and might 

 be considered due to hypertrophy and not atrophy ; but in nor- 

 mally lobed or compound leaves lateral protuberances would not 

 occur at all if the intermediate tissue kept pace with them, as is 

 the case with entire leaves. 



Lastly, two or more of the preceding causes may act conjointly 

 on a flower, inducing various modifications. Such is the case with 

 a monstrous Primula vulgaris communicated by Miss Dowson and 

 described by Dr. Masters in his paper " On some Points in the Mor- 

 phology of the Primulaceae" *, and which I have also examined. 

 Thus: 



Corolla. 



1. By atavism, the five petals are nearly free. 



2. By chorisis, the median fibro-vascular bundle bifurcates and 

 thus increases the number of lobes to ten. 



3. Solution has freed the corolla from the stamens, thereby 

 disproving the theory that the corolla o^ Primula is an outgrowth 

 from the androecium. 



Stamexs. 



1. The solution just mentioned separates them from the corolla 



* Now in the press for the Society's ' Transactions/ 





\ 





